Development and research of a novel seam tracking sensor based on active acoustic signal

Abstract The quality of weld seam tracking in automatic welding directly affects the welding quality and production efficiency. The weld seam tracking sensors commonly used in engineering are mainly based on laser vision and arc. These sensors are easily interfered with by factors such as arc light, spatter, and workpiece mirror reflection, which reduces the real-time tracking effect. Base on the above, in order to improve the accuracy and stability of weld seam tracking, a novel sensor based on an active acoustic signal is developed. The sensor detects differential changes in the active acoustic signal at various groove positions to identify and correct the welding torch’s position in real-time. The frequency response characteristics of the sensor are simulated and studied. The results show that sensor has the highest sensitivity to weld deviation detection when the frequency of the active acoustic signal is 14.4 kHz. The experimental results also verify the effectiveness of the research. The V-groove weld deviation experiment is conducted, and the mean amplitude, standard deviation, and short-time energy features of the active acoustic signal under 0-4 mm deviations are collected and extracted. The analysis shows that the standard deviation feature is the most sensitive to the deviation identification. The weld deviation identification model is established based on the standard deviation, and the V-groove weld seam tracking experiment is conducted. The results show that the deviation is within ±0.5 mm in 74% of cases and within ±0.9 mm in 93% of cases, indicating that the active acoustic signal sensor can achieve reliable tracking, providing a new solution for welding seam tracking.